Optimizing laboratory practices for recombinant adeno-associated viral vectors: impact of stress factors on vector stability

Abstract Recombinant adeno-associated viral vectors (rAAV) are currently considered the most promising platform for in vivo gene therapy. Several products have already achieved market approval, and many more are in the clinical pipelines, targeting a wide range of tissues and diseases. As we have recently reviewed, most rAAV products are still delivered in a deep-frozen state. However, efforts are emerging to enable storage of gene therapy products at higher temperatures. The main focus of these efforts can be grouped into vector optimization and formulation development. To efficiently screen formulation compositions, several laboratory-scale unit operations need to be taken into account. These operations require freeze–thaw cycles, handling at room temperature, and filtration steps introducing interfacial stress, all of which can potentially impact the quality attributes of rAAV vectors. In this article, we examine a number of critical unit operations that need to be conducted in order to screen rAAV formulations. These include freeze–thaw cycles, buffer exchange, sterile filtration, and in-use compatibility of clinically relevant rAAV vectors carrying a reporter gene. Regarding thermal stability, our focus is on short- to mid-term stability, which is applicable to the handling of rAAV samples in the laboratory, rather than on conditions for long-term storage. We included rAAV2, one of the most sensitive serotypes, as a ‘worst-case scenario’ for most stresses, as well as rAAV9 and engineered vectors. Our data show that formulations with neutral pH, high ionic strength, and the addition of sucrose have increased stability against thermal stress and multiple freeze–thaw cycles. Formulations at slightly acidic pH (pH 5.2) are beneficial for physico-chemical stability but have a detrimental effect on potency. Upon shaking, none of the commonly used surfactants (poloxamer-188, polysorbate-20, or -80) prevent aggregation. Therefore, preventing adsorption of the capsids to the contact materials remains the main purpose of including a surfactant in rAAV formulations. Buffer exchange can introduce significant stress to rAAV vectors in a serotype-dependent manner. Dialysis was shown to be the safest method to prevent titer loss and aggregation during buffer exchange, due to the minimal shear stress it introduces. Lastly, we found that sterile filtration does not significantly impact the quality attributes of rAAVs, taking rAAV8-based capsids as a representative example. In conclusion, we believe that the findings presented and the overall experience we share in this work will support the handling of rAAV vectors during research and development activities. Graphical Abstract